Version 3.5.0 release candidate
Pre-release
Pre-release
What's Changed
- Override NormalizedError function with CUDA by @sjsprecious in #424
- Auto-format code changes by @github-actions in #435
- 393 create cudasparsematrix class by @mwaxmonsky in #426
- Auto-format code changes by @github-actions in #437
- Optimize data movement for CudaRosenbrocksolver and compute minus Jacobian matrix on GPU by @sjsprecious in #436
- Auto-format code changes by @github-actions in #438
- Update the CUDA implementation of AddforcingTerms with the CudaVectorMatrix class by @sjsprecious in #439
- Auto-format code changes by @github-actions in #441
- Allow custom rate parameters to be directly set by @K20shores in #440
- Auto-format code changes by @github-actions in #444
- Rename CUDA matrix class to be Dense or Sparse by @sjsprecious in #443
- Auto-format code changes by @github-actions in #445
- Update the CUDA implementation of SubtractJacobianTerms with the CUDA matrix class by @sjsprecious in #446
- Auto-format code changes by @github-actions in #447
- add support for non-double properties by @mattldawson in #448
- Initial attempt to format cuda code. by @mwaxmonsky in #449
- Fixing bugs in previous PR. by @mwaxmonsky in #450
- Update template arguments for LuDecomposition by @mattldawson in #454
Full Changelog: v.3.4.0...v3.5.0-release-candidate